Skip to main content

Full text of "Proposal for a Research and Development Program on Computer systems"

See other formats

------------------------------<page break>-----------------------------
FOR PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
permission of tha insti:ute Archives 
Robert M. Fano 
Principal Investigator 
F. Leroy Foster 
Division of Sponsored Research 
Carl F. Floe 
Vice President, Research Admin. 
Charles H. Townes 
------------------------------<page break>-----------------------------
Dermission of tha institute Archives 
Appendix 1 - Research Areas of Current Interest 14 
Appendix 2 - Report 0f the Long Range Computation Study Group 
Appendix 3 - Faculty and Research Staff 
Appendix 4 - Pertinent Current Research 
Appendix 5 - Subjects of Instruction. 
------------------------------<page break>-----------------------------
FOR PERSONA USE ONLY. No part may be so'ld, :oanea copied, or pub]:shed w ''-" 
', . lnot the expre 
perission of tka institute Archives -!.I.?. 
1. National Computer Needs 
.The nation is facing urgent problems, bo'th military and civilian, in 
which conclusions must be drawn from, and timely actions must be taken on 
the basis of very large volumes of data collected from a variety of sources. 
Problems of this type are involved, for instance, in military command and 
control, air-traffic control, industrial stock and production control, and 
infomation retrieval for military, scientific, and health purposes. 
The information-processing aspects of these problems have been 
particularly frustrating. The initial wave of optimism, resulting from the 
apparently unlimited logical power of digital computers, was followed by the 
realization that much more was involved than the construction of larger and 
faster machines. Similar comments can be made about the use of computers 
in research. 
Computer technology has been progressing by leaps and bounds over the' 
last decade, in terms of reliability, size and cost of components, speed. of 
operation, and ease of assembly. On the other hand , the development of 
techniques for exploiting computers in non-numerical information processing, 
and'as aids in research and in human problem solving and decision making has 
been relatively lagging. Specifically, computer systems (including progrmmming 
aids, operational organization, and input, output, and display equipment) have 
not yet been developed that are easily and economically accessible, and that 
are truly flexible and responsive to individual needs, particularly the need 
for quick, direct response. There is substantial evidence that such computer 
systems can be developed, and tht'tWtl play a much more effective role 
in supporting and extending hum an cognitive capabilities. 
------------------------------<page break>-----------------------------
'Ol ?SRSONAt USE ONLY. No part ma?' be soid, .oaned, copied, or pub3[shed wlnmou the expre> 
permission of tha institute Archives - q.I.".- 2 - 
An "on line" mode of operation in which the individual scientist, problem 
solver, or decision maker is tightly coupled with a computer system of very 
large memory and speed appears attractive. It appears even more attractive 
as we envision the evolution of such a system to provide ready communication 
with others through machine information retrieval, including the development 
and use of open data files and public subroutines. On the other hand, in 
order for any such system to be economically feasible, the machine's memory 
and processing capacity must be shared simultaneously and independently  
by many on-line users in such a way as to insure its continuous, efficient 
exploitation. General purpose, independent, on-line use of computers by a 
large number of people has not yet been achieved, but it appears feasible on 
the basis of recent experiments with time sharing of large machines. 
2. The role of Universities 
The development of computer systems of the type envisioned equires the 
active participation of people engaged in research, problem solving, and 
decision making in many areas, to insure that the system will satisfy a 
variety of significant special requirements. In particular, the experience 
and suggestions arising from a cooperating community of imaginative users 
will be critical to the sound evolution of programming aids and terminal 
It ' is clear that universities can provide an ideal environment for such 
a development effort, because of the great variety of research and other 
pertinent activities that naturally flourish on their campuses Furthermore, 
universities are excellent test communities for experimenting with the notion 
of computer systems as public facilities, and in particular with the 
organization and use of public files and subroutines. Last, and most 
important, the students educated in t-me mzdt of such a research and develop- 
ment effort would provide the/necessary manpower for effectively attacking 
the information processing Rcoblems facing the Nation° 
It is equally clear, on the other hand, that no university can provide, 
or should attempt to assemble, the manpower and the production facilities 
necessary to engineer and mnufacture computer systems of the size envisioned. 
While university-built computers have been milestones in the growth of 
digital technology, the very special and limited resources of universities 
------------------------------<page break>-----------------------------
'OR PERSONAL USE ONLY. No pars may Oe sold, .oaned, copied, or published ''ihou she expre. 
permission of the nstiue Archives 
- 3 - 
should be focussed in the future on less developed aspects of computer systems. 
Similarly, the most effective long-term use of these resources in the solution 
of military problems, is their investment in the generation of new knowledge 
and techniques through work on scientific and other problems appropriate to 
3. Character of the effort needed 
The computer systems envisioned above should not be regarded as static 
aggregates of "hardware" and "software" designed and built to meet pre- 
determined requirements. Rather, they should be pictured as being in 
continuous evolution through addition, modification and substitution of parts 
in response to user experience, new requirements, and new technical develop- 
ments. As a matter of fact, the users themselves may be regarded as part of 
the system since their own approach to research and problem solving will have 
to evolve together with the hardware and software. 
Thus, a major research and development effort is required which should 
eventually involve most major universities and computer manufacturers, and 
last until computer systems are ready to become intrinsic parts of the 
organizations they serve° The ultimate goal, as vague as it may be, is 
-clearer than the road leading to it, and the initial steps are likely to be 
uncertain, and probably controversial. It seems, therefore, advisable to start 
with a pilot-plant operation at one university, working in close cooperation 
with a computer manufacturer, for the purpose of generating the experimental 
▀ evidence necessary to plan a broader-based effort. 
The close cooperation between a university and a computer manufacturer 
appears essential to the success of the effort, and yet fraught with legal 
and ethical problems. These problems will have to be explored experimentally, 
just as the technical ones, in th or,open minded cooperation. 
The Massachusetts Institute of Technology has been a leader in the field 
of computers and informatio?/processing since the pioneering work of 
Dr. Vannevar Bush on analog6e computers. It is now ready to spearhead a new 
research and development effort in the same field, by undertaking at once the 
program outlined below. 
------------------------------<page break>-----------------------------
FOR PERSONAL USE ONLY, No part may be sold, .oaned, copied, or published witout The expre> 
permission of th institute Archives '.I.I '" 
- 4 
1. Broad ob.]ectives 
The broad, long-term objective of the program is the evolutionary 
development of a cmputer system easily and independently accessible to a 
large number of people, and truly flexible and responsive to individual needs. 
An essential part of this objective is the development of improved input, 
output and display equipment, of programming aids, of public files and 
subroutines, and of the overall operational organization of the system. 
The initial development effort should be regarded as a pilot-plant 
operation intended to provide the knowledge and the experience necessary to 
plan a broader program involving an increasing number of universities and other 
A second, concomitant objective is the fuller exploitation of computers am 
aids to research and education, through the promotion of closer man-machine 
interaction. This second objective is not only important by itself, hut it is 
also essential to the development of the computer system envisioned above, and 
vice versa. 
The third Objective, which must be part of any university activity, is the 
long range 'development of national manpower assets through education in the 
oertinent area; of Faculty as well as of students, and outside M.I.T. as well 
as within the confines of the campus. Again, this third objective is 
inextricably interwoven with the preceding two, because people's approach to 
problems will have to evolve in parallel with the computer hardware and 
2. Initial computer system  
The development of the cmputer system envisioned above will be by 
necessity a sort of bootstrap operation. In fact, the evolving system will be, 
simultaneously, the main reJearch tool and the primary object of experimentation, 
as well as the tangible product of the development effort. Thus, the point of 
departuremust be an existing computer system, capable of evolving in the 
desired direction during the initial phase of'the program. 
------------------------------<page break>-----------------------------
.FiCE: Tnzs maerzal may be protecec oy cop)'rzga ia:, (T˝e i7 .S. Codejr FHIS 
FOR PERSONAL USE ONLY. No part may be sold, oaned, copied, or published without the expres 
permission of the institute Archives I.I.T. 
The system most appropriate for this purpose is the one presently 
available in the M.I.T. Computation Center. It consists of an IBM 7090 
computer with provisions for time sharing between four users: three 
on-line users operating through flexowriter ypewrtters, and one passive 
user of the Center's Monitor System. The number of possible on-line 
users will be increased to 16 by March 1963, and plans have been made 
for connecting to the system the IBM 1620 computer of the Civil Engineering 
Department and the PDP-1 computer of the Electrical Engineering Department, 
to be used as on-line consoles. 
Unfortunately, the Computation Center installation is already overloaded, 
and cannot supply the cmputer time required by the proposed program. 
Furthermore, experimentation with the system would be drastically limited by 
the obligation on the part of the Computation Cent,r to provide continuous 
service to a large number of users. Thus, it is planned to acquire, as part 
of the proposed program, a computer system which would in essence duplicate 
the Computation Center installation, except for the substitution of a 7094 
for the 7090 machine° 
3. Tentative Program 
The character of the long-term program requires that immediate attention 
be paid to the first three successive steps involved, namely the evolution 
and exploitation of the initial system, the aquisition of the basic 
hardware and software for a computer system of substantially greater 
capacity and speed, and the evolution of such a system in terms of 
operational organization/programming aids, terminal equipment, etc.o.The. 
following activities are concerned, in order, with these successive steps. 
3.1 Work on initial system. Presentplans for research and develop- 
ment on and with the initial' computer system amoun in essence to extending, 
speeding up, and integratiDg work already in progress'. With regard to the 
evolution of the initial system in terms of terminal equipment, the 
following three projects apear attractive at this time: 
a) The PDP-1 computer of the Electrical Engineering Department is 
a machine designed specifically for on-line operation, and its scope-light pen 
combination permits very close and convenient man-machine interaction. On 
the other hand, although its internal operating speed is high, its overall 
------------------------------<page break>-----------------------------
,FiCE: Yals material may be proecec oy cop)'rgh 
FOR PERSONAL USE ONLY. No part may be sold, .oaned, copied, or pub]ished without the expre$ 
ermission of the institute Archives 
- 6 - 
processing capacity is rather modest. %his suggests using this computer as a 
console, time-sharing the much greater processing capacity of a 7090 or 7094 
computer installation. 
b) The Civil Engineering Department is presently using an IBM 1620 
computer in conjunction with a graphical display unit, as an aid in solving 
various geometric problems arising in the design of roads. This installation,' 
li"e the PDP-1 installation, provides convenient man-machine interaction but 
it has a limited processing capacity. Again, valuable experience could be 
gained from using it as a console, time sharing a much larger computer. 
c) ihe Electronic System Laboratory and the Mechanical Engineering 
Department hae been cooperating in the development of techniques for machine- 
aided design. As ?art of this effort a dimplay console has been designed 
possessing a nu:'aber of novel eatures intended to facilitate the generation 
and manipulation of graphical information. The actual construction of this 
console and its conectin to the initial system would provide a terminal 
eminently suited for man-machine graphical communication. 
These specific projects are parts of research efforts in progress or 
of current interest, which are particularly pertinent to the objectives of 
this proposal. Some of these research efforts, together with their objectives 
and motivations are briefly discussed in Appendix 1. 
3.2 Acquisition of larer system. The ability of the proposed initial 
system to evolve in the desired direction is severly restricted by the 
inheret speed, memory,and structural limitations of the 7094 computer. While 
much valuable experience and information can e gained from the initial system, 
its organization and overall capacity appear zo be inadequate for meeting the 
objectives of this proposal. Thus, a machine of much greater capacity must be 
procured at the earliest possible date. 
The report of the Longange Computation Study Group of M.I.T.(dated 
April, 1961) attached to this proposal as Appendix 2, outlines the specifications 
of a computer system capabl'e of meeting most of }.I.T.'s needs a few years 
hence, and discusses various procurement alternatives. This report, and the 
feasibility of the computer system recommended, were discussed with 
representatives of several computer manufacturers during the late Spring and 
Summer 1961. The general reaction was that the computer system envisioned was 
technically feasible, and could be expected to be operative in about three years. 
------------------------------<page break>-----------------------------
,.:FCE: This material may be pro[ecse 
FOR PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the expre$,. 
permission of the institute Archives - .!.I.".- 7 - 
While this report is somewhat outdated and keyed [o a narrower objective 
that those of this proposal, its conclusions are still very. pertinent to 
the proposed acquisition of a suitable machine. 
Having a computer built to M.I.T.'s specifications still appears to 
be unwise for a variety of reasons, some of which are discussed in the above 
report. A much more desirable course of action would be to establish a 
working relation with a computer manufacturer which would make it possible 
to discuss current machine designs, with [he objective of acquiring a 
prototype, perhaps with minor modifications, of a future machine. The legal 
and ethical issues involved in such a close working relation would have to 
be explored in great detail beforehand. A possible procedure for selecting 
a suitable computer manufacturer might be through the award of two or three 
study contracts of 3-4 months_duration. 
It is proposed at this time that a meeting, including some half a dozen 
people from outside M.I.T., be held as soon as possible for the purpose of 
discussing the report of the Long Range Computation Study Group, and making 
recommendations with regard to the acquisition of a machine of considerably 
greater capacity than the 7094 of the initial system. The meeting should 
last only two or three days, and be reconvened later on, if necessary. 
3.3 Study grouDs. The procurement of a machine of conptderably greater 
capacity is only part of the next development stage. The planning of terminal 
consoles, programming aids, and operational organization should also begin as 
soon as possible. For this purpose, it is proposed to hold through the Spring 
and Summer 1963 a series of relatively small and short meetings on these 
aspects of computer system development, in addition to the meeting concerned 
with the procurement of the machine itself (see 3.2)° The series should 
terminate with a larger two-week meeti_gith the objective of reviewing the 
conclusions reached'in 
M.I.T.and A.R.P.A. with regar to the whole program. 
Each meeting will include people from other universities and from 
industrial and governmental'laboratories as well as from M.I.T. Position 
papers will be prepared and distributed ahead of time as bases for discussion. 
------------------------------<page break>-----------------------------
FOR PERSONAL USE ONLY. No part may be sold, .oaned, copied, or pub3ished without the expres 
permi&$ion of tha institute Archives - .!.I.T._ 8 - 
The following tentative list of meeting topics indicates the scope and 
purpose of the series: 
a) Meetings on computer system design 
Machine structure and procurement 
Input-output equipment and display consoles 
System organization and operation 
Programming languages and aids 
b) Meeting on user's requirements 
Artificial intelligence and experimental mathematics 
Psychology and biophysics 
Engineering design 
Economics, management and decision making 
Earth. sciences 
Language processing 
Education, including teaching machines 
c) Program review meeting 
Two-week meeting conce=ned with the three main objectives 
of the program, namely computer system development, 
research, and education° 
------------------------------<page break>-----------------------------
,orlC5: This material may be protected by copyright law (Title 17 U.S. Code). THIS C)?Y : 
FOR PERSONAL USE ONLY. No part may be sold, oBned, copied, or published without the expre 
permission of thg institute Archives - .!.I.T. 
1. Management 
The proposed research and development program will require the active 
participation of members of the Faculty and Risearch Staff from many M.I.T. 
Departments. and Research Laboratories. Thus, the program will be managed, 
at least initially, as a special project, operating largely through existing 
M.I.T. organization. 
The project will be directed by Dr. Robert M. Fano, Ford Professor of 
Engineering. In addition, because of the likely impact of this project on 
the Institute as a whole, a policy committee has been set up to represent 
and advise the Institute on matters concerned with the proposed program, and 
to advise and support Professor Fano in carrying it out. The committee 
repDrts to the Provost, Dr. Charles H. Townes, and is presently constituted 
as follows: 
Professor P.M. Morse, Director of the Computatio n Center, Chairman 
Dean G. S. Brown, School of Engineering 
Professor Po Elias, Head of the Department of Electrical Engineering 
Professor R. M. Fano, ex-officto 
Dean G. R. Harrison, School of Science 
Professor A. G. Hill, Physics Department 
Dean t{. W. Johnson, School of Industrial Management 
Professor C. Fo Overhage, Director of Lincoln Laboratory 
2. Research Staff 
It is envisioned that the program will eventually involve a rather 
large'umber of people from various mca emeprtments and Research 
Laboratories, as indicated b/he pertinent current research efforts' listed 
in ppendix 4. The members/of .th Faculty and R?search Staff listed 'in 
Appendix 3 have already agreed to devote a substantial fraction of thei 
efforts to the proposed program, and have actively participated in the 
preparation of this proposal. In addition, it is expected.that some 
20 graduate research assistants will be involved by Fall 1963. 
------------------------------<page break>-----------------------------
FICE: This material may be protected by copyright-law (Title 17 U.S. Code). THIS COPY l,. 
PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
rmission3of a.q9%eiR.Weves - .1.I.'ť. 
It is estimated that the project will require by Fall 1963 additional 
space for computer installation and staff of the order of 30,000 sq. ft. 
Various possible locations have been investigated on and in the immediate 
vicinity of the campus. The only suitable space btainable by Fall 1963 
is located at 545 Technology Square, in.a commercial office building 
scheduled for occupancy in the next few months. The top two floors of the 
building, amounting to approximately 30,000 sq ft., have been leased for 
20 years by C.E.I.R., which is apparently interested in subletting them 
for a period of 3-5 years. The terms of such a sublease are presently being 
------------------------------<page break>-----------------------------
ŞR ERSONAL USE ONLY. No part may be sold, %oanled _, copied, or pub3isked without tke expre$.- 
e.rmi$sion of tha institute Archives - 
1. Experience with on-line operation 
A central objective of the proposed program is to provide flexible, 
on-line access to computers. Computers specifically designed for flexible, 
on-line use have been available at M.I.T. for a number of years; the TX-O 
computer, built by Lincoln Laboratory, and presently located in the Research 
Laboratory of Electronics; the TX-2 computer, built by, and located at 
Lincoln Laboratory, and the PDP-1 computer, a gift of the Digital Equipment 
Corporation to'the Electrical Engineering Department, also located in the 
Research Laboratory of Electronics. By now, many students, as well as 
members of the research and academic staff have had the opportunity to use 
these computers in their work, thereby gaining experience with their 
exploitation and value as research tools. 
2. Lon5 R..ane Computation Study Group 
In late Spring, 1960, a committee was appointed by President Stratton, 
composed of Profeas0rs Fano, Hill, Morse and Wiesner, with Professor Hill as 
Chairman, for the purpose of formulating a policy with respect to the develop- 
ment of future computer facilities at M.I.T. A Study Group was, in turn, 
rganized to estimate the future computer needs of the M.IT community, and 
to recommend specific means for meeting them. The report of the Long Range 
Computation Study Group, published in April, 1961, is attached to this 
proposal as Appendix 2. The conclusions and recommendations of the Study 
Group were reviewed and accepted by the parent Committee. 
The main conclusion of the Stud?Group was that "The major part of 
M.I.T.'s computational needs, a fgw years rom now, can best be met by a 
single very high-speed larg pacity computer system with provisions for 
time-sharing through a number of remote consoles" and that "this system wold 
be the cheapest way of proiding the needed capacity and the only way of 
providing the necessary, close man-machine interrelationship that we feel is 
vital for research in the years to come." This conclusion and the system study 
on which it is based are at the roots of this proposal. 
3. Experimentation with time sharing 
Experimentation wih ime sharing has been in progress at he M.I.T. 
Computation Center for several years. The system presently in operation, 
* Fo Jo Corbato, M. Merwin-Daggett, R. C. Daley, "An Experimental Time-Sharing 
System," Proc. of the Spring Joint Computer Conference, May 1962. 
------------------------------<page break>-----------------------------
FOR ?5RSONAL USE ONLY. No part may be sold, .oaned, copied, or published without the expres 
permission of the institute Archives - .!.I.T. 
- 12 - 
as part of the Center's 7090 installation, makes provisions for four 
independent users. Three of the users can operate on-line through typewriters, 
while the fourth one consists of programs runniJg in 'the background through 
the standard Monitor System. The installation of an additional memory of 
3.2,000 words for storing the time-sharing supervisor, together with other 
equipment changes, will, by Spring 1963, increase the number of possible 
independent users to 16. 
A time-sharing system has also been designed for the Electrical Engineering 
Department PDP-1 computer, and'will be in oeration by Spring, 1963. This 
system has provisions for time-sharing eight typewriters, two scopes and light- 
pens, two tape readers, and two tape punches, as well as one or more real-time 
control computations through the computer sequence-break system Initial 
oper'ation will be with two typewriters, one scope and light pen, and one tape 
reader and punch. 
4. The role of computers in research and education  
The role of computers in research and education at M.I.T. is discussed in 
detail-in Sec. B(pp.14-26) of the report of the Long Range Computation Study 
Group, attached to this proposal as Appendix 2. The survey of computer usage 
on which this discussion is based was conducted during the second half of 1960. 
in early December, 1962, a questionnaire was sent to some 60 members of 
the Faculty and of the senior Research Staff to determine their interest in 
the objectives of this proposal, and the extent and nature of current work 
pertinent to them. The list of pertinent current research attached to this 
proposal as Appendix 4 was compiled from their replies° Because of the 
informality of the survey and of .kaftime-!mitations involved, this list is 
bound to be incomplete, and somewhat inaccurate. Yet, it provides a clear 
indication of the strong and support that the proposed program can 
be expected to elicit from the M.I.T. community. 
* J. Eo Yates, "A Time-Sharing System for the PDP-1 Computer," Report 
AFCRL-62-519 (MIT Report ESL-R-140), June 1962, OTS/ASTIA No. AD-285851. 
------------------------------<page break>-----------------------------
?OR PBRSONAL USE ONLY. No part may be sold, loaned, copied, or published witkout the expres 
permission of the institute Archives !.I.T 13 
5. Educational program 
A significant indication of the extent to which computer and computer 
related disciplines have spread their roots at M.I.To can be found in their 
rapidly increasing incorporation into the subjects of instruction offered 
thrpughout the Institute. 
'A simple measure of this phenomenon is provided by the curves in 
Appendix 5, obtained by counting the subjects whose description in the 
M.I.To. Catalogue either mentions computers explicitly, or mentions related 
disciplines, such as system analysis and information theory, but without 
reference to computers. The selection criteria employed are subjective but 
consistent throughout. Thus, the general shapes of the curves are more 
meaningful than the actual values° The subjects selected from the 1962-63 
catalogue are listed in the same appendix to illustrate the criteria employed. 
The total number of instructurs in charge of these subjects is 93, which 
amounts to approximately il °/o of the M.I.T. Faculty° 
------------------------------<page break>-----------------------------
permission of ths institute Archives .!.I.'7. 
CODled , 
- 14 - 
or pubJlshed without the expre 
Appendix ! - Research Areas of Crrent Interest 
Programming Languages 
An essential feature of anyman-machine problem solving system is 
a powerful flexible, and expressive language facility. There are a 
great many specialized programming ianguages in use today, but most of 
them are quite restrictive) both in areas of application and ease of 
expression. hile it is possible to conceive of creating a !age time- 
sharing cnputer system by merely providing a built-in executive routine 
and scheduling system so that each of the separate programming systems 
can operate independently and unchanged) such an approach is highly 
inefficient and poses mauy problems of compatibility. Furthermore, 
such an approach offers little in the exploitation of the unique 
capabilities for new problem-solving techniques which real-time, 
time-sharing makes possible. 
Recent adVances in the theoretical foundations of language mechanisms 
indicate that agorithms can be written which will permit the use of 
specialized progrsmning languages whose richness approaches that of 
natural lamoage. The physica form of the languages is of little 
importance -- the same algorithms will translate verbal or graphical 
!3_aes of almost any form. The incorporation of semantics as well as 
syntax into the theoretical structure also willpermit generglized 
operators to be used) so that such features as symbol manipulation can 
be included in a natural wy. The language algorithms are high/jr efficien - 
and.well-suited to multi-processor, mlti-user time-sharing system 
operations. Sinceacuser ..'ld up his own vocabula-y) and incorpor: 
in his own fashion, any problem-statement and symbol manipulation features 
he requires, a system built in this way will have much greater capability 
than is presently available. Further theoretical research, with re- 
duction of these new developments to practice) is expected to yield 
working systems for general use in the MIT time-sharing project as well 
as in computer-aided design. 
Comp.utr-aided design 
Computers have been used for some 5Lme in the engineering field for 
performing calculations difficult or repetitive enough to warrant the 
programming time required. There is considerable interest and effort at 
------------------------------<page break>-----------------------------
FOR ?ERSONAL USE ONLY. No par may be 
permission of tha institute Archives - M.I.V. 
- 15 - Aip. 1-a 
the present time in elevating the status of the computer "rcm a "glorified 
desk calculator" to an active real-time partner to a m=mu in solving 
types of engineering problems. For example, work is now underway to eter 
the Automatically Programmed Tool System (a problem-oriented !ang:age used 
off-line to translate engineering drawings into a form suitable for 
autcmatic calculation of machine-tool tapes for numerical manufacturing) 
into a Crampurer-Aided Design System, wherein the designer would use the 
cronpurer on-line to develop and refine a] . the specifications of a mecbni 
part or assembly, including .shape, weight strength, choice of standard 
catalog parts, etc. In the civil engineering field, there is interest in 
smilar extensions of problem-oriented languages now used for solving 
geometric problems in surveying, design of lighwys, etc. However, a majc 
lmtation in cmnputers which has slowed such applications to date has bee 
the lack of convenient and flexible means for on-line man-machine 
cation, particularly in the area of graphical and picture languages,. which 
are a pr !e form for engineering applications. It now appears 
that the oscilloscope and light-pen combination can be further developed 
to provide the necessary facility. Particular problems to be solved in 
designing remote and time-shared consoles for design use include: faster 
flicker-free disp!ays inclusion of character and geometric figure 
generation independent of the main computer, to free up computing time and 
memory, and ease data-link requirements  and convenient and flexible means 
for control by means of light-pens, switches, dials, and alpha-numeric 
System Programming 
To develop for.'y particulc9mputer  general purpose time-sharing 
system with the ,ritical time-sharing supervisor program a number of 
specific system programming areas must be investigated and appropriate 
techniques explored and developed. These progrg techniques are 
intimately connected with the hardware and should properly be included in 
the cemputer design. Critical areas to be investigated are: 
1. The interrupting and switching frcm one program to another 
within the high-speed memory. 
2. The swapping of programs in and out of the high-speed memory. 
3. Memory protection and relocation including the data transfers 
of input-output equipment. 
------------------------------<page break>-----------------------------
,,.riC5: ?ns material may be prozec:ea o>' cop?'rlghz law [TIz˝e i7 J,S. Code); FHIS C?l : 
FOR PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the expres 
permission of tha instit:ute Archives - :.].I.?. - 16 - App. 1-b 
4. The subroutine linkage of user programs with the supervisor, 
especially in terms of verification of !egitte parameters. 
5- The ab.ty to write read-only subroutines for simultaneous 
use by many user programs. 
6. The appropriateness of time-of-day clocks accounting clocks 
and interrupt clocks. 
I. The ability to handle input-output easily within the supervisor 
including the assignment and control of devices such as tape 
units, scopes, disc memories, drum memories. 
Other system programming areas which must be developed are: 
1. A control language format includir easy. input typing 
. General utility inspection programs such as post-mortems, 
traces differential dumps etc., where the user should be 
able to specify information symbolically.  
3. Inter-console and inter-program cammunication conwentions. 
4. Coordination techniques of all messages in and out of the user 
Information and program storage and retrieval conventions. 
Accounting and auditing procedures for monitoring system usage. 
In the area of reliability there must be programmed rovision for: 
1. The systematic investigation of suspected main computer ma3anctic 
without unnecessai 7 system shutdown. 
2. Notification proceduresf0r system filures and shutdowns. 
3- Devel9Ument of a gradationof backup procedures and calculated 
losses for the 'inevitable mishaps which can occur to any 
computer system. 
Conducting experiments with on-line ccmputation 
In many research applications the computer is used as a sophisticated 
data processing device in which the data sources are experiments being 
carried out in research laboratories. Scme salient examples are nuclear 
physics experiments with high energy particle accelerators, aect obser- 
vation of the behavior of animal nervous systems, and the evaluation of 
psychophysical experiments. Numerous other possibilities await adequate 
computation facilities for serious exploitation. 
------------------------------<page break>-----------------------------
,,.FiCE: This material may be proecec o)  copp'rlgn law l.]'le i7 O.S. Code- I'HiS CPI : 
FOR PERSONAL USE ONLY. No part may be sold, oaned, copied, or published without the expres 
permission of th instiute Archives - M.I.T. - 17 - App, 1-c 
At present, it is genera3_ly necessary to record data during the 
exeriment, and process it later when computer time is ava3_lable. Not onl 
does this require expensive equipment or tedious transcription, but the 
delay in processing and presenting results is frustrating to the experi- 
menter. Frequently, the processed data, if immediately available would 
enable the experimenter to change the conditions of the experiment to obta 
more significant data. In many cases the computer can even serve to contr 
the conduct of an experiment with more accuracy and precision than the 
human experimenter, and with more flexibility d sophistication than 
devices he could afford the time and effort to construct. 
In scme instances, smali general purpose ccmputers are used to proces 
such data on-line. However, the character of the processing is quite 
limited by the capabilities of the equipment being used. One objective of 
the proposed research will be to determine how,the capabilities of a 
powerful processor can be made available for on-line experimental 
applications through time-shared operation. 
Research in computer-drected instruction 
For the past three years a program of research in autcmatic 
instruction has been carried on at MIT. The purpose of this program has 
been to develop a teacking machine which has as many of the desirable 
characteristics of the humantutor as possible. The program was motivated 
by the feeling that existing teaching machines did not serve as tutors but 
rather as drillmasters who seem to bore rather thanto stimulate the 
average student. Inordero achieve the sophisticated decision making 
properties of Se human tutor, tleast partial, it has been necessary 
to use a logics& element with powerof a genera&-purpose cemputer. Con- 
sequently, the research has focussed o how a teaching machine governed by 
a ccmputer could be made an efficient and stimulating participant in the 
The difficulty of most previous machines was that the program of 
instruction could not be sufficiently adapted to the capabilities and 
interests of the student. A student who answered a given question incor- 
rectly a second time was required to repeat the same sequence of instructiŻ 
It ws decided that one of the design goals for the cmputer-rected 
mackine would be to have it present to the student at each stage in a 
teaching process, instructive materis& suited to his state of knowledge 
in the subject and to his innate ability. 
------------------------------<page break>-----------------------------
permission of thz institute Archives M.ť. ' 
Such a computer-directed teaching machine has been built using the 
Itt 7090 cputer at MIT. The output device of the teacb_ing machine is a 
microfilm reader modJ_fied so as to move to a particular frame upon cd 
of the comptrter. The machine has displayed an ability both to vary its 
presentation of course material depending upon the past perfformance of eac: 
studnt and to change its decision. logic as a result of the past per- 
formance of ] stttents. Preliminary studies indicate that as many as 
%.'U'0C stylents could be taught different subjects by the computer at the 
sme ie if an appropriate system of remote teaching console:-: were 
operated in a time-s,haring mode. %_is type of oilration appears to be 
economic even at the presen time. 
Yu its ufttimate development the computer-frected teachir machine 
may function not only as an individa] tutor but as the control_ling 
element for many of the phases of education. For example, the eaching 
system may assign outside work for the student, grade his performance on 
this outside work, and adapt its future outside assignments to the 
capabilities of the student for unsupervised instruction. 
Research on a useful "mathematical assistant" rogrm 
Once the time-sharing system is in operation, we believe we can make 
available to the scientific world a new kind of service -- the Mathematical 
Assistant Program. This wl be a service for cooined synfooliŻ and 
numerical computation. At the user's request the program will apply 
selected transformations and manipulations to mathematical expressions 
inserted at the consola.--The ,er' reqies't 'm z'-e frc simple substi- 
tution of variables through expansions in power series Fourier series u. 
other rather c.mplex mathematical operations; to applying complex solution 
metis for . solviug systems of 8ferential equations, solving 
eigenvaiue problems, etc. Of course everything the syst em does must be 
based on progrm written into the system, and much work wl]. be required 
in the construction of the programs themselves, and the devising of a 
szltable control .!anagge. But the effect wo'] be worth many times the 
effort we believe. It should open the way to solving more complex 
nalic problems (in'physics and engineering as well as in matheanatics) 
than can be attempted today. It can also sere greatly to increase the 
mathematical power of the scientists using the system. For in many cases 
the scientist will need only to know the conditions appropriate for applyir 
------------------------------<page break>-----------------------------
oermi$sion of thg institute Archives "I ' 
- 19 - 
App. 1-e 
a certain method; he may not need to develop the technicst proficiency 
now required to competentl perform a mathematical operation. It is our 
view that ultimately the mathematical assistant may become one of the 
most important developments in scientific methodology opening a form of 
mathematical competence to scientists in all fields. 
Research on heuristic .ro6rmn  
Heuristic progrmug is the construction of computer programs to 
solve problems using sophisticated methods of exploration, learning, 
pattern-recognition, and other information-processing techniques. 
has-already an outstanding record of achievement in this area. The 
availability of time-sharing.methods should make it possible to extend 
this research into areas which have up till now, presented forbidding 
complications. The reason is that it was necessary in the past to program 
the entire process before one could begin to explore the qus_lity of its 
performance. This means that it was necessary to write programs for 
making every decision that the machine might encounter. With the time- 
sharing system it will be feasible to run programs that contain lacun 
with the missing decision-making capability supplied by the console 
operator. In this way we can build up complex programs by steps where 
in the past we were presented with the task of working out in advance an 
entire structure of forbidding ccmplexity. From the viewpoint of appli- 
cation as wll, the prospects of cnplex aids to problem-solvug seem 
much more attractive in the new framework. For example, w have developed 
a reasonably effective..prmrerformng some of the problem tasks 
connected with Integral Calculus. 0nthe other hand, without time-sharing 
any practical use of such a program requires a very large and special 
effort. We expect to find ediate practical use for the program as' 
soon as the time-sharing system is operating on a large scale. 
Operational studies 
The trend in large cnputer systems is toward greater complexity in 
design and greater par_11 elisin in operation. No large amount of fore- 
sight is required to envision computer systems whose memories are shared 
by several processors and whose processors are used frc numerous consoles 
These possibilities raise important and interesting operational problems 
not =le the problems found in a busy telephone system or a bustling 
------------------------------<page break>-----------------------------
zOTtChl This material may be protected by copyright law (Title 17 U.S. Code). THIS COPY ,. 
OR PERSONAL USE ONLY. No part may be sold, ioaned, copied, or pubJished without the express 
▀ e=maaion of the institute Archives - :.!.I.T. 
job shop. Research on sme of these problems is currently in progress at 
the School of Industrial Mamagent and at the Operations Research Center. 
Most attention to date has been on sequencing and priority problems. 
The priority problem arises when one or more classes of customers in a 
servicing operation merits sue for of preferential treatment. In a 
time-sharing computer system, for example, users differ in the urgency of 
their requests how long they have been waiting for service, and how much 
additional time they require. These differences should be reflected in 
the procedures adopted for assigning priorities to the users. The question 
is how?. 
By way of illustration, short requests for time should take precedence 
over lomger requests for time, other things being equml. This contributes 
to a high overall service efficiency; that is, to low average delay. In 
practice however atending to this objective may cause partially con- 
flicting objectives to suffer. One study underway attempts to balance 
the conflicting objectives with the help of priority models, phrasedin 
mathematical terms and with economic considerations incorporated. 
Management information and decision sMstems 
Computer oriented research at MIT in the area of mauagement and 
decision making typicyhas divided into five main headings. 
l. Basic research into process-oriented languages and computer 
system design; also, cmpiler development. 
2. Simulation studies the levels. of individual firm, industry 
and econO. 
3. Business games in the areas of\,marketing, production, and 
4. Statistical studies, some of which employ a back-and-forth 
inter]playwiththe computer. 
5. Operations research: analysis of sequencing, priority, 
mathemgticalprogrmng, and other decision models. 
Recently, there also has been a great amount of interest in the 
development of future computer systems and their application directly to 
the management process particularly to top-level decision making. 
Envisioned are management information systems which provide top management' 
with timely fimancial inventory, marketing and production data with no 
------------------------------<page break>-----------------------------
:,O'FiC5: This material may be protected by copyright law (Ti:le 17 0.S. Code). THIS COPY 
FOR PSRSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
parmission of th institute Archives - 
- 21- App. 1-g 
more disccfort or inconvenience than that associated with the placing 
of a telephone cadl. Also envisioned are management decision systems 
whose central elements in sxtdition to the management decision makers 
themselves are c=puter simulations and mathematical decision models 
run from time-sharing czputer consoles. Several simulation studies 
currently in progress have this objective as their primary focus. 
Smb olic processes 
It has been observed that the wy a person approaches a problem is 
strongly influenced by the lang-=ge available to hm -- a well-known 
example is the approach to multiplication or division using Rgn or 
Arabic numerals. Thus the facilities available in a progrsmmng lang__sge 
may be crucial to the successful utilization of a machine. Besides 
appropriateness a progrsmng language must also offer convenience, 
brevity fluency legibility ease of learning and ease of checkout. 
Much of the naturalmess of some progrmm!ug languages comes Crcm certain 
intuitively recognizable underlying smlarities to English, yet a ro- 
grng language must be much more restricted than English partly because 
of our severly limited ability to mechanize English. These restrictions 
cause problems to the user, who must be aware of the limltations, yet 
not too seriously confined by them. 
Research is in progress in the theory of progrmmug and into the 
practical engineering ccmprcmises required of successful 
Longer ruge research is in p-regess in the mechanical translation 
group in the areas of the syntax and Semantics of nglish, logic, and 
the philosophy of laneo The  of this research is to develop a 
capability i a cmputer for handling English menug that is for 
understanding unrestricted English. 
------------------------------<page break>-----------------------------
O'FICE: This material may be protected by copyright law {Title 17 U.S. Code), THIS COPY 
FOR PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the expres 
parmiaaion of tha institute Archives !.I.'f. 
- 22 - 
 Research in machine structure 
Presently available machines are not oriented in their structural 
design toward multi-user, remote console operation. The relation between 
processing units and memory,.the means of coupling typewriters and display 
devices to the machine, and the provision of'data terminals for real-time 
interaction with laboratory experiments are subjects that require the 
evaluation of new ideas. Many concepts of machine organization can be 
properly evaluated only through operational experience: One must determine 
how rsearch workers and students utilize and react to a system designed 
with new concepts, and one must measure the degree to which system components 
are efficiently used in typical circumstances. 
For example, a computer designed with several processing units, each 
having access to a common pool of modular memory is veryattractive as the 
core of a multi-user system: It promises to provide efficient utilization 
of a large memory, greaer reliability, ease of expansion, and the 
possibility of temporarily assigning the full use of one processor to a 
particular user when justified by special needs. The evaluation of this 
C concept requires the construction of such a system and its utilization in 
an academic environment. 
------------------------------<page break>-----------------------------
O'rtCE: This material may be protected by copyright law (Title i7 U.S. Code). THIS COPY i. 
FOB PťRSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
permission of th institute Archive5 
Appendix 3 
The following members of the Faculty and Research Staff intend to devote 
a substa.uial part of their efforts to the proposed program 
CORBATO, Professor Fernando J(ose), Department of Electrical Engineering 
and Computation Center, Massachusett Institute of Technology, 
Cambridge 39, Massachusetts. PHYSICS. Oakland, California, 
July 1, 26, m. 62. B.S. California Institute of Technology, 50; 
UoS.N. 44-46; M.I.T. Computation Center: Research Asociate, 1956-59; 
Assistant Director in charge of Programming Research, 1959-60; 
Associate Director, 1960-. Associate Professor, Department of 
Electrical Engineering, 1962-o Tau Beta Pi; Sigma Xi; Physical 
Society; Asno Computing Macho, Editorial Board, 1962-; computer 
programming and applications; molecular and solid state physics. 
Professor Corbato's principal professional interest is in the 
techniques of computer usage. He is presently active in the 
development of time-sharing systems. 
Professor Jack B(onnell), Department of Electrical Engineering, 
Massachusetts Institute of Technology, Cambridge, Massachusetts. 
October 13, 1931, Elizabeth, New Jersey. SoBo, S.M., 1954; 
SCoD., 1958 (Electrical Engineering) M.IoT, Teaching Assistant, 
1954, Research Amsistant 1955, Instructor 1958, Asistant 
Professor 1959 M.IoT. Fields of interest: Speech communications 
researchl Computer structure for experimental computation; 
Mathematical programming. 
Professor Robert M(ario), Ford Professor of Engineering , 
Department of Electrical Engineering, Massachusetts Institute of 
Technology, Cambridge, Massahusettso ELECTRICAL ENGINEERING° 
Torino, Italy, Nov° tl, 17; nat; m.49; c 3. SoB. Mass° Inst. of 
Tech. 41, Sc.D. (eleco eng.s$t. elec. eng Mass. Inst. of 
Tech. 41-43, instro 43-44 radiatiohlb. 44-46, res. assoc., res° 
lab. of electronics, .46-47. asst. prof.XEleco Communications, 
47-51; assoc prof.,s51-56, Prof. 56-61, 9rd Prof. of Eng. 1961-;- 
group leader Lincoln Lab. 51-53, Fel. Inst. Radio Eng; Fel. Am. 
Acad. Microwave circuit components; network synthesis; transmission 
of information; theoretical limitations on the broad band matching 
of arbitrary impedances. Author: Transmission of Information; 
Electromagnetic Fields Energy, and Forces (with L. Jo Chu and 
R. Bo Adler); Electromagnetic Energy Transmission and Radiation 
(with R. B. Adler and Lo Jo ChU) o 
------------------------------<page break>-----------------------------
O'FICE: This material may be protected by copyright law (Title i7 U.S. Code). THIS COPY' i.. 
FOR PERSONAh USE ONLY. No part may be sold, loaned, copied, or published without the express 
pe:mission of th institute Archives - I.I.T. 
App .3-a 
Dr. Frederick C., A.B. DePauw Univ. 1938, Ph.D.(Psychology) 
Columbia Univ. 1947. Served in London, OWl, and USAAF during war. 
Lecturer at Columbia Univ. 1946-47; Instructor,Harvard 1947-40. 
In 1950 made Chief, Communications Research Div., USAF Human 
Factors Operations Research Laboratory; in 1954-55 served as 
Acting Director, Electronics Research Directorate, USAF Cambridge 
Research Center. Joined MIT Lincoln Laboratory 1955 as Leader 
of Psychology Group, appointed Associate Head, Information 
Processing Division 195'9. Appointed Division Head in 1960. For 
the past few years his major research interest has been in the 
area of automatic speech recognition. 
Professor Martin, School of Industrial Management, Massachusett 
'Institute of Technology, 50 Memorial Drive, Cambridge'39, Mass.. 
APPLIED MATHEMATICS. Elizabeth, New Jersey, Nov. 30, 31; m. 59. 
A.B., Harvard, 55, Nat. Sci. Found. fel. 55-56, A.M., 56, 
Ph.D. (Applied Mathematics), 58. Mgr. appl. sci., Int. Buso 
Mach. Corp., 57-58; asst. prof. Indust. Mgt. Mass. Inst. of 
Tech., 58-61, Assoc. Prof., 61-; Consult., opers. eval. group, 
U.S. Dept. Navy, 60-; USAF, 52-54, Capt. Opers. Res. Sot.; 
Ash. Computing Mach.; Inst.Mgt.Sci. Application of the computer 
and mathematical methods to economic; industrial nd management 
problems; simulation of economic systems; quantitative methods, 
'Notes on a new pseudo-random number generator'(j.Asn.C0mputťn $ 
Mach.); co-auth., 'icroanalysis of Socioeconomic Systems"; 
ed, "Management and the Computer of the Future." 
Professor David A(lbert), Department of Electrical Engineering, 
Massachusetts Institute of Technology, Cambridge, Mass. ELECTRICAL 
ENGINEERING. Alliance, Ohio, Aug. 9, 25. B.E.E., Ohio State., 44,. 
M.S.E. 49; Sc.D. Mass. Inst. Tech., 53. Instr. elec. eng. and res. 
assoc., univ. res. found., Ohio__State, 46-50; asst. pof.Elec. Eng., 
Mass.. Inst. Tech., 50-57c-rof., 57-62, Prof., 62-;.Levy 
medal, Franklin Inst., 55.. USNR, 446. Inst. Radio Eng. Fellow  
62-; consult. for President's Sci. AdWComm. 1959-; Air Force Sci. 
Adv. Board 1959-62. $Wic...kl'B7crutť:echhiqe; aomata. and 
information theory. 
Professor Marvin L(ee), Department of Electrical Engineering 
Massachusetts Institute of Technology, Cambridge, MasS. New York 
Cityi.'Aug. 9, 27. B.A. Harvard 1950, M.S Princeton, Ph.D. 
Princeton, 1954 (Mathematics). Junior Fellow, Harvard Soc. FelloWs 
1954-57, Staff Member Lincoln Lab. 1957-58, Asst. Prof. Math. 58-6i  
Assoc. Prof. Elec. Eng. 62-; Staff member RLE,MIT Computation Center. 
Director MIT Artificial Intelligence Group. Editor: J.A.C,M. Membe= 
I.R.E., AMA, AAS, NYAS, Sigma Xi, Areas of Interest: Heuristic 
Programming, Artificial Intelligence', theories and brain. function;'.. 
recursive function theory, mathematical logic, epistemdlogyo 
------------------------------<page break>-----------------------------
ii0TLC5 This material may be protected by copyright law (Title 17 U.S. Code). THIS COPY i' 
F0 PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
permission of the institute Archives - ?.I I.T. 
▀ ' App. 3-b 
Douglas T(aylor), Electronic Systems Laboratory, Massachusetts 
Institute of Technology, Cambridge, Mass. MATHEMATICS. Canton, China, 
Dec. 21, 29, m. 51o A.B. Cum Laude, Oberlin, 51; S.M., M.I.T., 54. 
Lecturer, E.E.M.I.T., 60-; Head, Computer Applications Group, 
Electronic Systems Lab., M.I.T. 1956-; Div. of Sponsored Res. Staff, 
M.I.T., 1952-. Project Engineer, Computer-aided Design Project, 1959-; 
Technical Director of following projects: PRE-B-58 Fire Control System 
Evaluation Programming, 1952-56; Automatically Programmed Tools (APT) 
System, 1956-1959; Computer-aided Design, 1959-. Outstanding Young Man 
of Greater Boston Award, 1958. Sigma Xi, Assoc. for Computing Machinery, 
AAAS. Language theory, programming systems, computation techniques, 
man-machine communications, numerical control of production processes, 
artificial intelligence, problem-solving systems. 
Oliver G., Lincoln Laboratory, Massachusetts Institute of Technology, 
Lexington, Massachusetts. MATHEMATICS, London, U.K., May 10, 26. 
S.Bo M.IoT. 45. Staff, Lincoln Lab., M.I.T. 52-. Consult. for 
PSAC 59-° Artificial Intell., Computation Techniques, Automata 
SIMPSON, Dr. Stephen M(ilton) Jr., Department of Geology and Geophysics, 
Massachusetts Institute of Technology, Cambridge, Mass. GEOPHYSICS. 
New York, NoY., Jan. 29, 29; m. 62° B.So, Yale, 50; Ph.D., M.I.T. 53.  
Instr. geophysics, MoI.T. 53-54, asst. prof. 54-62. Computer scientist 
RCA, 57; assoco prof., geophysics, Mass. Inst. fech., 62-. Dir. MIT 
Vela Uniform project 60-. Phi Beta Kappa, Sigma Xi, er. Geoph. Union. 
Geophysics, seismology, time series analysis, digital computer 
programming, weapon system control. 
Professor Herbert M., ELECTRICAL ENGINEERING, Canton, Ohio, March 20,30; 
m. 53; c. 2, S.Bo, MIT, 52, National Sci. Found fellow, 52-53, 
Sc.D. (Elec. Eng.), 55. Asst. Elec. Eng., MIT, 53-54, instr., 54-56, 
Asst. Prof., 59-. Consult., UoS. Navy Bu. of Ships, 59-; Curriss- 
Wright Corp°, 69-; Inst. Defo Analysis, 61-; American Optical Co., 
61-; USNR, 55-59, Lt. DigitleOmlapplication and design, 
partzcu!arly man-machzne systems, large s.vtems, and connrol. 
Dr. John E(rwin), Electronic Systems Laboratory,XMassachusetts Institute 
of Technology, Cambridge, Mass° ELECTRICAL ENGINEERING. Toledo, Ohio, 
Jan° 4, 20; mo 49, B.S., MIT, 43; M.S., MIT 47. Research staff, MIT 
Radiation Laboratory 43-45. Research ASSto, MIT, 45-47° Staff Member, 
MIT Div. Sponsored Research, 47-56. Res. Elec. Engineer, MIT, 195-. 
Project Engineer, Servomechanisms Lab., MIT 49-55° Executive Officer, 
Servomechanisms Lab., MIT 55-59. Assistant Director, Electronic 
Systems Laboratory, MIT, 59-. Sigma Xi, Inst. Radio Eng., Assoc. of 
Comput. Mech. Auto. Control, digital computer applications, radar, 
weapon fire control, data reduction, instrumentation, man-machine 
------------------------------<page break>-----------------------------
OTICE:. This material may be protected by copyriRh 
FOR. PERSONAL USE ONLY. No part may bŞ o, oand c  >ubiahd without the expres 
permission of ths institute Archives - 
Dro Victor H(use), Department of Electrical Engineering, Massachusetts 
Institute of Technology, Cambridge, Mass.' SYMBOLIC PROCESSES. 
Niagra Falls, N.Y., July 5, 20, m.43, c. 3; B.S. Antioch College, 43, 
S.M. Univ. Chicago, 1951, Ph.D. (Physics) 1953. Research Asst., 
cosmic ray research 1947-53, staff member Res. Lab. of Elec., 1953-54, 
Asst. Prof. Modern Languages 1954-61, Research Assoc. Elec. Eng. 1961-. 
Director Mechanical Translation Group 1953-. Consultant, National Bu. 
of Standards; Editor, Mechanical Translation; President, Assoc. for 
Machine Translation and Computational Linguistics; Linguistic Society 
of America, Assoc. for Computing Machinery; American. Standards Assoc. 
Areas of interest: Mechanical translation; Design and construction 
of programming languages; theory of programming languages; linguistics; 
machines that understand; information retrieval. 
------------------------------<page break>-----------------------------
iO'PlCg: This material may be protected by copyright law (Title i7 U.S. Code). THiS 
FOb PSRSONAL USE ONLY. No part may be sold, loaned, copied, or pubJished without th 
pezmiaaion of tha institute Archives - .!.I.T. 
Appendix 4 
List of pertinent current research compiled from replies to a questionnaire 
sent to approximately sixty members of the faculty and research staff. 
Feasibility study of photographic storage techniques for digital 
information and weighted area scanning techniques for optical character 
recognition. Part of a continuing project aimed at developing automatic 
libraries for scientific and technical igformation. A hi,h-speed, h/gh-densty 
photographic storage device has been built and another version is nearin 
completion. An optical character recognition scheme has been simlated om the 
7090 and a full-scale reading machine is being desined. 
Operation of on-line computer for control systems experiments on 
neurological systems. Pattern classification studies. Measurement of 
estimation of constraints on random processes. 
Setting up a New England Center for computer and information technology in 
the biomedical sciences. 
BIOPHYSICAL COMMUNICATIONS (Siebert, W,Rosenblith, W., Morse, H., Peake, W. 
Goldstein, M., Gerstein, M., Eden, M., Weiss, T. 
Correlation and statistical analyses of electrophysiological data, 
especially discrete events from single .neural--unis. Development of special 
/ converters.Simulation of well 
purpose analog and digital computers./and 
defined pieces of neural physiological system on the TXl,2, especially. 
peripheral parts of the auditory system. Computer recognition of handwriting. 
COGNITIVE PROCESSES (Torgerson, W. S., Yntema, B. B., White, B. 'W.) 
Research in the area of complex judgement and decision processes. 
COGO (COordinate GeOmetry) is a programming language and processor for 
solving eometric problems on a compute r. Applicable to computational 
problems in contour surveys, highway design, right-of-way surveys, inter - 
change design, bridge geometry, subdivision design construction, layout, etc.. 
System is currently being extended and developed as a research tool for 
machine communication. 
------------------------------<page break>-----------------------------
:OTICS:. This material may ba protected by copyright la (Tiule 17 U.S. Code). THIS COPY 
FOR. PSRSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express'. 
permission of the institute Archives - M.I.T. 
Several studies directed toward improving communication with the bind 
including design of a high-speed electric Brailler and revision of the Braille 
rules to improve information transfer efficiency. 
COMPUTER-AIDED DESIGN (Ross, D., Ward, J., Coons, S., McCltntock, F., 
Kurtz, E., Baumann, D. M. B., Mann, R.) 
Investigation of means for facilitating the creative activities in an 
engineering design process by effecting a symbiosis with the computer. Direct 
and intimate interaction of man and computer in concept-development and 
Exploration of management decision processes whose central elements are 
computer simulations and mathematical decision models run from time-sharing 
Construction of device and programs to accept handwritten 
comptero Development of time-sharing equipment on IBM 7090. 
To study individual and group decision-making'behavior under partial 
uncertainty conflict. Experimental environments have been programmed on 
computer, which allow systematic variation and laboratory control of 
stimulus availability, response or alternatives availability, reward 
structure, complexity of situation, etc. 
Developing a new design philoshy for computer stems based on lists 
and declarative representation of algorithms. Emphasis on man-machine 
integration, interaction betWeen independent computers. 
GRAPHICAL DISPLAYS ON-LINE (Roberts, L., Gladding, E., Foster, R.) 
Development of plotting system on-line with 1620 'computer. 'Application 
to presentation of computer-generated design information. Use for display of 
engineering information upon which design decisions Can be made. 
Heuristic program to solve symbolic integration problems.. Currently 
writing heuristic program to obtain explicit solutions to differential equations. 
Several other programs under development. 
------------------------------<page break>-----------------------------
.i0'ricE: This material may be protected '07' copvrzgh iu'.; !,Ti:ie , J.S. Code) /Mils 
FOR-P. 5.RSONAL USE ONLY. No part may be sold, ':oaned, copied, or published without the express 
pa=miasion of ths institute Archives - ,'.l.I.':'. 
App .. 4-b 
HUFAN INFORMATION PROCESSES (Torgerson, W.. S., Harris, W. P., Mitchell, R. T., 
Schulman, A. I. Smith, J. E. K., Stowe, A. N., 
White B. Wo, Yntema, B. B., Forgie, J.) 
Automatic speech recognition and speaker identification by computers. 
General research on complex judgement and decision processes (identification 
evaluation, memory, and perception) with computer simulation of these 
processes as their rules are discovered. Man.-machine cooperation in decision 
Study by computer simulation of underlying structure and behavior of firms, 
indusries, and economies. Emphasis on interaction of separate elements and on 
flowa of men, materials, money capital equipment, ordera, and information. 
Study of structure and requirements of information systems in management 
enterprises government, and the military° 
Civil Engineering Department developing a laboratory facility for student 
instruction and research in information systems. Being equipped with wide 
range of electrical, mechanical, optical, and electronic information handling 
components. Present small computer being provided with facility to communicate 
with large number of different kinds of input/output devices. 
LINGUISTICS STUDIES (Chomsky, N., Halle, M., Mathews, H., Stevens, K.) 
Study of restricted infinite automata (bounded memory growth machines, 
more structured than general Turing machines). Incorporation of grammars into 
the domain of such machines. Speech Souhd generationsing an electroacoustical 
analog of the vocal tract driven by/computer commands formed from inputs 
approximating word spellings. Computer verification of and experimentation with 
hypothetical grammars. 
LOGICAL DESIGN (Susskind, A.) 
Work on logical design of computers, special purpose computer design, 
analog-digital conversion techniques, and pulse-circuit design. 
------------------------------<page break>-----------------------------
pc=mission of thz institute-Archives b!..'ť. App'.. 4-c 
Research in methods of mechanizing English and other languages. Study of 
limited temporary memory on language structure. Imprevement and standardization 
of programming languages (COMIT). Basic research on language and how it carries 
Simulation of individual decision-making. processes, especially as they 
pertain to economic behavior, using the Complex Information Processing approach 
of Simon and Newell° 
Research on models of human operator as a control element and as a decision 
maker in various display-control situations. Also research in programmed 
PATTERN RECOGNITION (Selfridge, 0., Neisser, Uo, Minsky,Mo) 
Pattern recognition and learning with respect to computer programs. 
Optimization of programs which assimilate large quantities of data, organize 
it and search for and recognize significant features in ito Applications to 
visual recognition problems and to chess° Analysis of nerve networks. 
PDP-1 COMPUTER (Dennis, J.) :. 
Time-shared operation for education and research° Work on hardware, 
consoles, programming, and compilers. 
PRIORITY PROBLEM (Greenberger, Mo) 
In a time-sharing system, the next user to be served is selected on the 
basis of nature of request, urgency, and time waiting. Such considerations 
typically lead to conflicts° This s.uRyanes the resolution of these 
conflicts through t e analysis of mat ematical models 'of priority systems, 
and the formulation of economic objectives. 
Research in the patterns of technical information° Experimental model 
of a Technical Information System now under development. 
Processing of psychophyical data, preparation of punched tapes for 
control of displaym on psychophysical experiments, and computer mtudy of 
models of sensory communications systems. Remearch on problems related to 
sensory aids for blind. Use of computers in unusual picture processing 
preparatory to coding of visual information for tranmmision to sense 
modalities whose capacities are much mmaller than that of visual ense. 
------------------------------<page break>-----------------------------
pezmisaion of the institute Archiyea - ..'7. 
ipp. 4-d 
Simulations of election campaign strategies, media-mix reach and frequency, 
and conication in communist bloc. 
Use of computer to operate a vocal tract model to synthesize speech, 
requiring man-machine and machine-equipment iBteraction. 
Research program to develop a language and programming system for analysis 
and design of structures. Will enable the user to specify in common 
engineering terms the shape, makeup, and properties of a structure, the loads 
and other conditions to be investigated, and the type of answers to be produced. 
Research in automatic instruction decision systems; particularly those 
systems that use a computer as the decision element. 
3D INFORMATION STORAGE AND RETRIEVAL (Loeb, A., Robinson, W., Haughton, E.) 
Use of algorithms from discrete analytical geometry for the storage, 
retrieval, and interaction computations of crystal structures. Optimal 
coding of more general 3-dimensional concepts. Program controlled manipulatin 
of 3-dimensional models applied to concept training in human8 by teaching 
TIME SERIES ANALYSIS (Simpson, S., Galbraith, J., Claerbout, J.) 
High speed storage and retrieval of long time series with associated 
descriptive material. Programming techniques for ultra-fast computations of 
correlation functions, spectra, minimu-p asor-d spectra, prediction 
operators, optimum filters, etc. Compilers for automaic subroutine testing 
and documentation. 
VOCODER RESEARCH (Gold, B., Rosen, P.) 
Research and development on VOCODERS (banR compression devices for 
speech analysis, band compression, transmission and resynthesis) utilizing 
special purpose digital computers for determination of pitch and employing 
digital transmission. Simulation of process on general purpose computers. 
------------------------------<page break>-----------------------------
OK FiKhUNAL US ONLY. NO part may De SOlU, ioane, cope oF- published without the express 
permission of the institute Archives - >i.I.'?. 
Appendix 5 
List of One-term MIT Subjects whose 1962-63 Catalogue Descriptions Involve: 
X Information Theory Control Theory, Systems Analysis 
XX Specific Mention of Computers or Data Processing Machines 
X 1.03T Social and Political Factors in Engineering 
XX 1.04 Mathematical Methods in Civil Engineering 
X 1.07, 1.08 Engineering I, II 
XX 1.15 Computer Approaches o Engineering Problems 
1.16J Special Studies in Information Systems (A) 
XX 1.19 Engineering of Location (A) 
X 1.21 Fundamentals of Transportation Systems (A) . 
XX 1,212 Transportation Systems Analysis (A) 
XX 1.251 Theory of Traffic Flow (A) 
XX 1.252 Design of Urban Highway Systems (A) 
X 1.26 Ground Facilities for Air Transportation (A) 
XX 1.503 Naval Structural Analysis 
XX 1.571, 1.572 Advanced Structural Analysis (A) 
XX '1.591 Numerical Methods of Structural Analysis (A) 
XX 1.951 Construction Management (A) 
XX 1.952 Construction Management Senar (A) 
XX 2.00 Introduction to Engineering 
XX 2°03 Dynamics 
XX 2.215 Methods of Engineering Analysis (A) 
XX 2.722 Descriptive Geometry 
XX 2.723 Elementary Nomography 
XX 2.724 Nomography 
XX 2.731 Engineering Design 
XX 2°732 Design Concepts 
XX 2.735 Man-Machine Systems 
XX 2.736 Moderh Kinematic Problems 
Note: (A) Following subject title indicates 
graduate subject. 
Instructor in charge 
C. L. Miller 
M. J. Holley 
P. O. Roberts, 
C. L. Miller 
C. L. Miller, 
H. M. Paynter, 
W. W. Seifert, 
Y. T. Li 
P. O. Roberts 
A. J. Bone, 
A. J. Bone. 
A. J..Bone 
R. F. Hansen 
J o J. Connor 
So J. Fenyes 
A. G. Dietz 
Ao G. Dietz 
H. M. Paynter 
S H. Crandall 
S H. Crandall 
D P. Adams 
D P. Adams 
D P. Adams 
R W. Mann 
D. M. Baumann, 
H. H. Richardson 
T. B. Sheridan 
Do P o Adams 
------------------------------<page break>-----------------------------
FOR, .5RSONAL OSE ONLY. Ng par% may be sold, gand, copied, or published without the 
parmimaion of .th institute Archives - '4.I.':'. App, 5-a 
2.745J Special Studies in Information Systems (A) 
XX. 2o751 Analysis and Design of Engineerin s Systems I(A) 
XX 2.752 Analysis and Design of Engineering Systems II(A) 
XX 2.78 Hydraulic and Pneumatic Controls 
XX 2.782 Introduction to Automatic Control 
XX 2.783 Advanced Automatic Control (A) 
XX 2°784 Istruments for Measurement and Control (A) 
XX 2.859 Manufacturing Technology 
XX 4.74 Methods of Research in City Planning (A) 
X 6.05, 6.053 Circuits, Signals, and Systems 
X 6.213 Feedback Control Principles 
X 6.214 Feedback Control Laboratory 
XX 6.216 Dynamics of Electric Machines (A, except for 
X 6.232 Electromechanical Control Components 
XX 6°25 Digital System Application 
XX 6.251 Digital Computer Programming Systems 
XX 6.252 Digital Systems Engineering 
6o27J Probabilistic Systems Analysis 
6.291 Radar-System Engineering 
6.311 Principles of Communicatfon 
6.36 Speech Communication 
6.37 Sensory Communication 
6.372 Introduction to' Neuroelectric Potentials 
XX 6.38 Pulse Circuit Analysis 
X 6.39 Image Transmission Systems 
XX 6.41 Introduction to.Automatic Computation 
XX 6.533 Algorithms for Optimization and Approximation (A) 
XX 6.536J Systems Engineering and Operations Research (A) 
XX 6.537 Switching Circuits (A) 
Instructor in charge 
H. M. Paynter, 
C. L. Miller, 
W. W. Seifert, 
Y. T. Li 
H. M. Paynter 
H. M. Paynter 
S. Y. Lee 
J. L. Shearer 
J. L. Shearer 
S. Y. Lee 
N.H. Cook 
A. Fleisher 
W. M. Siebert 
L. A. Gould 
L. A. Gould 
C. Kingsley : 
R. H. Frazier 
H. M. Teager 
F. J. Corbato 
A. K. Susskind, 
F. C. Hennie 
R. A. Howard 
J. F. Reinties 
J. Mo Wozencraft 
M. Halle, 
K. Stevens 
W. Rosenblith, 
S. J. Mason 
M. Goldstein, 
W Peake 
H J. Zimmermann 
W Schreiber 
F McCarthy 
A A. Goldstein 
R A. Howard 
D A. Huffman, 
E. Arthurs 
------------------------------<page break>-----------------------------
FOR. PERSONAL USE ONLY. No part may be sold, loaned, copied, or published without the express 
permission of the institute Archives 
App. 5-b 
Subject Instructor in charge 
XX 6.539 Mathematical Theory of Computation and 
Artificial Intelligence (A) 
X 6.54 Sampled and Quantized Systems (A) 
XX. 6.541J Language, Symbolic Processes, and Computers(A) 
XX 6.542J Mechanical Translation and Language 
Processing (A) 
XX 6.544J Heuristic Programming (A) 
X 6.571 Statistical Theory of Communication (A) 
X 6.572 Statistical Theory of Nonlinear Systems (A) 
X 6.573 Random Signals and Noise (A) 
X 6.574 
X 6.575 
X 6.576 
Transmission of Information (A) 
Advanced Topics in Information Theory (A) 
Statistical Theories of Signal Detection (A) 
X 6.59 Bioelectric Signals (A) 
XX 6.592 Analytical Models for Human Processing 
of Sensory Inputs (A) 
X 6.595J Biological and Neurological Control Systems(A) 
'X 6o601 Feedback Control Theory (A) 
X 6.602 Nonlinear Control Systems (A) 
XX 6.604J Special Studies in Information Systems (A) 
6.606 Control System Synthesig (A) 
6.607 Feedback Control Laboratory (A) 
6.608 Process Concrol (A) 
6.626 Modulation Theory and Systems (A) 
6.629 Radar System Engineering (RA) 
7.593J Biological and Neurological Control Systems(A) 
7.86 Molecular Genetics (A) 
7.97, 7.98 Molecular Biology Seminar (A) 
8.75 Operations Research, Principles and 
Applications (A) 
F. McCarthy 
R. A. Howard 
V. H. Yngve 
V. H. Yngve 
M. L. Minsky 
Y. W. Lee 
Y. W. Lee 
W. B. Davenport, 
W. M. Siebert 
R. M. Fano 
P. Elias 
W. M. Siebert, 
W. B. Davenport 
W. Rosenblith 
M. Eden 
L. Stark 
L. A. Gould 
W. W. Seifert 
W. W. Seifert, 
C. L. Miller, 
H. M. Paynter 
Y. T. Li 
G. C. Newton 
G. C. Newton 
L. A. Gould 
J. Mi Wozendraft 
J. F. Reintjes 
L. Stark 
C. Levinthai 
F..O. Schmitt, 
P. F. Davison 
P.M. Morse, 
'H. P. Galliher, 
R. A. Howard 
------------------------------<page break>-----------------------------
FUR ERU'L%h S2' ONLY. No pars may Oa so. ic, ioaned, copied., or pub] ishe.& without the express 
permission of the institute Archives - 
Appg. 5-c 
Instructor in charge 
xx 10.34 
XX. 10.35 
X 10.55 
X 12.831 
Numerical Solution of Problems in 
Chemical Engineering (A) 
Dynamics and Control of Chemical. 
Engineering Processes 
Economic Balance in Chemical Industry(A) 
Elements of Geophysics I 
XX 12.851, 12.852 Geophysical Analysis and Computation 
XX 13o03 Advanced Hydromechanics of Ship Design 
XX 13.96 Application of Operational Methods 
XX 14.11 Linear and Nonlinear Programming 
X 14.532 Goals and Strategies in Foreign Affairs 
X 14.781 Seminar in Learning (A) 
X 14.79T Psychology of Language and Communication 
X 14.791T Seminar in Psychology of Language and 
Communication (A) 
X 14953 Mass Media and Communication Systems 
X 14.954 Methods of Communication Research (A) 
X 14.956 Public Opinion and Propaganda 
X 14.96 Power, Influence and Policy Decision8 
X 14.771 Seminar on Sensation and Perception (A) 
X 15.04 Organization 
X 15.092 Mathematics for Industrial Management \, 
XX 15.094J Probabilistic Systems Analysis 
X 15.095 Statistical Decision Theory (A) 
X 15.12 Management Problems I 
X 15.51 Industrial Accounting and Control 
XX 15.542 Management Information Systems (A) 
XX 15.58 Industrial Dynamics 
XX 15.581 Industrial Dynamics I(A) 
XX 15.582 Industrial Dynamics II(A) 
X 15.592 Mathematical Programming(A 
X 15.593 Stochastic Systems (A) 
XX 15.594J Systems Engineering and Operations 
Research (A) 
P. L. Brian 
P. L. Brian 
G. C. Williams 
T. Cantwell, 
T. Madden 
S. M. Simpson 
M. A. Abkowitz 
E.G. Frankel 
R. M. Solow 
L. Po Bloomfield 
D. H. Howes 
D. H, Howes 
I. Pool 
F. W. Frey 
I. Pool 
F. W. Frey 
J. A. Swets, 
D. M. Green 
M. Greenberger 
R. A. Howard 
 R. A. Howard 
R. A. Howard 
H. Hudgins 
J. C. Emery 
J. C. Emery 
E. B. Roberts 
E. B. Roberts 
J. W. Forrester 
M, Greenberger 
------------------------------<page break>-----------------------------
FOR, ?ERSONAL USE ONLY. No part may 5e so, io&ne, co..e&, es pub]ished without the express 
permission of the institute Archives - 
15.599 Research Seminar in Quantitative 
Analysis (A) 
15.71 Production Management 
X. 15o72 Industrial Standardization 
X 15.791 Operational Models of the Firm (A) 
X 15.792 Manufacturing Decisions Seminar (A) 
XX 15.82 Marketing Management 
XX 15.821 Marketing Policy and Administration (A) 
X 15.83 Marketing Research 
XX 15o832 Demand Analysis and Marketing Research(A) 
X 15.92 
X 16.30 
X 16.31 
X 16.32 
XX 16.33 
XX 16.362J Special Studies in Information Systems(A) 
App. 5-d 
Instructor in charge 
M. Greenberger 
E. H. Bowman, 
B. E. Goetz 
W. Schreiber 
E. H. Bowman 
H. J. Claycamp 
G. B. Tallman 
R. B. Maffei 
R. B. Maffei, 
R. M. Cunningham 
Organization (A) - - 
Principles of Automatic Control W. McKay 
Principles of Instrumentation and Control(A) Y. T. Li 
Principles Of Instrumentation and Control(A) Y. T. Li 
Automatic Control Systems Lab.(A) Y.T. Li 
Y. T. Li, 
C. L. Miller, 
X 16o37 
X 16.38 Physical Components of Control Systems (A) 
X 16.40 Control Systems Principles 
X 16.41T Control System Engineering)  
X 16.42 Weapons Systems (RA) '/ 
X 16.43 
X 16.44 
X 16.46 
XX 16.74T 
Statistical Problems in Automatic Control(A) 
Weapons Systems Lab (RA) 
Automatic Control of Flight Vehicles (A) 
Astronautical Guidance (A) 
Advanced Flight Vehicle Engineering (A) 
16.761, 16.762 Orbital and Ballistic Flight (A) 
16.78 Control and Guidance in Flight 
Transportation (A) 
H. M. Paynter, 
W. W. Seifert 
W. W. Vender Velde 
R. K. Mueller 
Ho P. Whitaker 
J. Hovorka 
W. Wrigley 
W. R. Markey 
R. K. Mueller 
H. P. Whitaker 
R. H. Battin 
R. H. Miller, 
M. A. Hoffman 
P. E. Sandorff, 
M. A. Hoffman 
H. P. Whitaker 
------------------------------<page break>-----------------------------
FOR, PERSON'L USE ONLY. No part may bo sold, loaned, cupuu, u puujuu -uuuc u= u.p'.a 
permission of the institute Archives - .l.I.?. 
X 18.11 Operations Research (A) 
XX 18.163 Heuristic Programming and Artificial 
Intelligence (A) 
X 18.80 Topics in Foundations of Math.(A) 
X 19.34 Statistical Methods in Meteorology 
X 19.35 Statistical Problems in Meteorology 
XX 19.46 Numerical Weather Prediction (A) 
XX 19.67 Planetary Fluid Dynamics (A) 
X 22.241 Principles of Control Systems(A) 
XX 22.242 Nuclear Plant Dynamics (A) 
XX 22.26 Nuclear Reactor Design (A) 
XX 22.53T.Digital Computers in Nuclear Engineering(A) 
XX 22.531 Digital Computers in Nuclear Engineering(A) 
X 23.751 Language and Society 
XX 23.771 Mathematical'Backgrounds for Communications 
XX 23.772 Mathematical Models in Linguistics 
XX 23.791J Language, Symbolic Processes, and Computers 
]iX 23.792J Mechanical Translation and Language 
Processing (A) 
Instructor in charge 
G. P. Wadsworth 
M. L. Minsky 
H. Rogers 
E. N. Lorenz 
E. N. Lorenz 
N. A. Phillips 
J. G. Charney 
E. Gyftopoulos 
E. Gyftopoulos 
H. Fenech 
K. F. Hansen 
K. F. Hansen 
M. Halle 
B. Hall 
N. Chomsky 
V. H. Yngve 
V. H. Yngve 
------------------------------<page break>-----------------------------
permzsson o th instztute Archiv '-,1.I.%.' - 
------------------------------<page break>-----------------------------